INDUSTRIAL A N D ENGIhlEERING CHEMISTRY
596
tested for tensile strength and elongation. Curve I11 and Table I1 give the variation in physical properties with the variation in ammonium chloride number. Conclusion
From the above experiments it can be seen that there are a large number of factors affecting the physical properties
of rayon during the spinning process. The time of contact of the newly formed filament apparently is of importance and if carefully controlled should help in obtaining yarns
Vol. 22, KO. 6
of higher grade. Under present methods of spinning actual contact with the spin bath, a t relatively high temperatures, is very short-never more than 0.5 second. Complete regeneration is attained on the spool by the adhering spin liquor a t a relatively low temperature or by the action of heat, during the washing and drying processes. These processes are usually adjusted by experiment to give the maximum physical properties and might well 6e shortened and made more exact by an attainment of complete regeneration through control of time of contact with the spinning bath.
Radiant Heat from Radiant Heaters and Its Measurement’ F. E. Vandaveer .4YERICAS
T
GASA S S O c I A T l O N TESTING L A B O R A T O R Y , 1032 E A 5 r 62ND S T . , CLEVELAND, O H I O
H E value of r a d i a n t The thermopile and its calibration, the method of the “biologic” effect of ultrae n e r a , emitted by the testing radiant heaters for radiant efficiency used at violet rays is widely knomm, sun, for p r o l o n g i n g the American Gas Association Testing Laboratory it is not generally realized have been described in detail. h u m a n life and improving that visible energy and the It has been shown that: (1) the thermopile is capable shorter infra-red rays have a general health conditions is of measuring radiant heat as emitted by a radiant definite stimulative v a l u e , becoming universally recogheater; (2) the method of test is capable of giving an usually a t t r i b u t e d t o t h e nized. Medical science has accuracy within * l per cent of the true reading; and power they possess of peneshown that people who have been deprived of the benefits 13) the diathermacy constant obtained is consistent trating the outer layers of of exposure to the sun, either and may be accurately obtained. the skin. A higher intensity by reason of adverse climatic Several means of increasing the total radiant heat can be borne of incident enc o n d i t i o n s o r because of have been given, as well as a discussion of possibilities ergy of the visible or short of future research work along this line. being confined within doors, infra-red quality than of the long i n f r a - r e d , and this is are much more susceptible to diseases of various kinds such as tuberculosis, pneumonia, thought to be due to the greater penetrativeness of the neurasthenia, etc. Remarkable relief has been afforded chil- shorter rays. The energy emitted by the sun contains more dren suffering from rickets and malnutrition by giving them visible and short infra-red energy than that emitted by a sun baths and otherwise bringing them into more direct contact domestic fire, and this would appear to explain why it is with the beneficial rays of the sun. The widespread interest possible to bear in comfort more sun energy than of the enin this is reflected by many of the more prominent hospitals ergy produced by an ordinary red-hot body. These conand sanitariums in prescribing sun baths and exposure to siderations indicate a direction of possible improvement in artificial ultra-violet rays in the treatment of certain diseases. the gas radiant heater which should tremendously increase As a knowledge of its medicinal value increases, more and its value for heating and health purposes. more attention will be directed toward duplicating and Considerable research has been conducted along the aboye scientifically controlling this type of radiant energy by arti- lines in this country, but the published scientific information ficial means. The health-giving rays of the sun cannot pene- on the subject is meagre. Possible fields of research are mentrate fog, smoke, dust, ordinary window glass, or solid ob- tioned in two commercial booklets (9, I 6 ) , which, it is believed, jects of such material as are generally used for constructing should be encouraged and developed. houses and factories. In many large cities fog, smoke, or We are interested not only in the spectral quality of the heat dust obstructs the sun’s rays a great proportion of the time, emitted, but also in the quantity, or the efficiency of a heater particularly during the winter months, when the days are for producing radiant heat. It is proposed in this paper t o short and the inclination is to remain inside as much as give a brief review of published literature, to summarize the possible. Statistics show that during the winter months work done a t the Testing Laboratory of the American Gas there are a great many more deaths than during the summer. Association on measuring radiant heat, and to point out how Therefore, if this vital radiant energy is to be obtained, deaths the radiant efficiency of space heaters may be improved. decreased, and general health conditions improved, artificial Published Work radiant energy must be provided. Prior to 1910 few references are made to instruments suitable .i.ery logical Source of supply of such energy is from room heaters of the radiant type. Energy from gas-fired radiant for measuring radiant energy from gas heaters. In 1910 heaters approaches, to Some extent, that emitted by the Callendar (8) described a radiobalance, an ingenious device sun and, by application of the necessary research, could in which the radiometric receiver consisted of a thermoprobably be made to gi\re off a greater percentage of the type junction which could be heated electrically to neutralize the of energy desired. Considerable scientific work has already heat generated in the receiver by absorbing radiant e n e r a . been done in England (IO, IC). It is pointed out that, while Coblentz, of the U. S. Bureau of Standards, wrote several valuable papers ( 3 , 4,,5, ?) in 1913 to 1916, describing his 1 Received March 15, 1930 Presented before the Division of Gas On determining radiant The and Fuel Chemistry at the 70th Meeting of the American Chemical Society, by the iinierican Gas Association Testing Laboratory, and Atlanta, Ga., April 7 to 11, 1930.
deicribed i n this paper, was made according to his reconuneiidatimn, based on the information contained in these papers. Hone and collalvirators ( 1 ) describe a holornetric method of detmniiiing the efficiencies ofradiating bodies. Radiation Srwn the soiircc falls on two coils of platinurn wire wound ~ z p ma thin piece of mica for finnnes;r and coat.ed for const.anoy t o an even surface with hard black enatnel, t,lie two coils heiiig iiaiunted hack to back on either side of a circular guniiietal box provided with wattrr circulation arid suitable covers for t.he coils. The coils are connectwi to a gal\-anoinetcr for readings. To keep constant the zero of the indicator, the temperatiire of the receirer o u s t he Inaintained constant, :ind therein lies oiie diffimilty in its genrral iisc,.
p '.
time apprmimat,ely six liutrdred heaters hsvc been tested for radiant efficiency aid the results obtained have 1)eee consistent and romparatively accurate as will he shown latpr. Description of Thermopile and Smpport
The tbermopilc nsed for deti:rininatirrn of radiant, heat Enim gacfired radiant heaters is shown in Figure 1. It. consists of a cylindrical brass brrx, ;I,2 ) / 2 inches (6.4 ciii.) in diaiiieter tiy 2 inches (5 em.) in deptli painted on the outside with white enaiuel and on the inside with black eiiine el. h Iwas shield, B , also coat,etl witli white enainel, the elcetrir ronneetiorrs from direct heat, Emin t l r r heater. To regulate tlie area of tlie hot jiinotioos cxlxwxl to tlii? heat. manicled Iirass shutters, c3art: used. Tlie end o f shutter D is recessed inch (6 nim.) t,r, prevent. wiwectwl Ireat fmni affcrting the thrrniopile. I)irectly back of the slit a transparent quartz disk grouiirl and polislied to inclr (2.4 ciii.) dianreter and 0.025 inch (0.64 nun.) tliirk is lomted The wiring of the therinopile shown in the lowcr 1+14e consists of twelve coi~plesof ~.r)M'er-ciinstantan COIInected in series. At each junrtion the wires are soft-aklered t o a tin plate nmi. square and 0.022 inm. t,hick paint,e:l, on the front side only, vjitli a solihon of larnphlack in alrohol a i d then coated with the soot of a speriii candle. T ~ cPo p per wire. I . is 0.0015 iiicli (0.038 nini.) in diaiiiet,cr and tlip constaiitao wire, ,I, is 0.003 inrh (0.0ifi inni.) in iliarneter rolled flat to 0.02 inin. Tlie tin receivers, E , of the Iiot junc-
tions are in t l i e m d hut not elcctriral contact. this heing aca,rnplished by slrellacking the edges of t.lrti rcroivers and overlapping wlien asseinbling. Twelrc of the junctions, E, are arranged together directly back of the slit in t,he la,x, the other twelvc: jonc,tiiins, FMure 1-Thermopile for Defermlnaflon of Radlanf Neaf F , rritli six on cacti side of the hot jimt:tions, are so placed In 1923 Moll (18) described a iherinopile for measuring that no radiation can reach t,benr. Copper lead wires extend radiat.ion. The cold jiiiictions of the thcrmogile were in con- from the thennopile to t h e I,inding post,s, G. Approximately tact vith metal masses which kept. down their teinp~ratore. 3,J,8 inch ( 5 nini.) back of the thwinopile is a11 isiiiglas, K , In ordev that the Iiot, jiinctiims might have small heat ca- fastened to blie insulator, L, bo ellclose the thcrrnopile arid pacity: tlie himetallic strips ompo posing the thermopile were fnrther pmt,ect it from air currents. A steel rod, Is, to v k l i made of plates of constantan and maiiganin silver-soldered. :in apparatus for snpporting t,lie thermopile inay he attaclieil, This tliexiioyilr is similar to the inst.runierit being used a,t extends from the back of t h e iiistriiiiieiit.? this laboratory. .I radir, tliennonicter (1.1) has been used for measuring radiant heat Iry nobing tlie temperature rise of a weighed aiimunt of vater in a copper box situat,ed ahout 3 feet from the heativ. The box is insulat,ed o n all sides except bhat facing t,he h a t e r , this side heing hlackend t.o increase the absorption effect,. This irietfrod is open to bwo objections: (1) The heat absorbed may be affcctrd by air current,s passing over the rough blackeiieil surface of the hnx; ( 2 ) the t,inie reqiiired for ~neasuritmentof temperature rise of suitalrle iiiagniturle iii about, 1 hour. The intensity of radiation through the space in front of the heater varies widely a t differmt points, so that many rrieasurerrrents must he made. Reihr and Knoblauch (17) studied radiat,im from doiiiestic rooiii heaters by niralis of a bolrimeter and apparentl? olitainrd good results. Itadiatinn, limited (I::), of 1,ontlorr gives a general view of apparatus (onsrreened) used in ineasuring the penetraiiay of radiant energy. The apparatus consists of a sliecinirn Iiolder, spectrometer, lanip and holders, and galvanometer. In 1926, when research for approval reguireinent. corninittees of tlie American Gas Association was being conducted Fieure 2-Apparafus for Defermlnstlun of Radisnt Heat preparat.ory to writing roqirireinentii for radiant, hcaters, it became necessary t o devise an apparatus which would ~ I C B . " In use, the tlienrropile i8 nioimted 0x1 a support which tire radiant efficiency with comparative accuracy and at t,he same time he rapid enough for routine testing. Aft,er trying slides on B a//r-incli (1.9-cm.) bar foniied iii the arc of a circle some of tlie methods in use a t that time W. W. Cohlentz was of 2-ftmt (til-cn~.)radius. This slipport. is shown ill Figure 2; cnnsulted and as a result the tlicr~iiopilcdescribed by Shawn 3 This inrlrcmieiii l i e s designed and r r r d r by the lipples Labwatory, (19) and discussed in detail herein was rlrvrlnpcrl. Since that sewi,ort. n. I.
598
INDUSTRIAL AND ENGIAINEERING CHEMISTRY
The bar is mounted in a vertical position with pivots at top and bottom so it may be swung along the arc of a circle. The holder a t the top of the bar is graduated in 10-degree intervals and on the bar itself there are stops for the thermopile a t 5-degree intervals. Readings may be taken over an area equivalent to approximately two-thirds of a hemisphere of 2-foot (61-cm.) radius. Calibration of Thermopile To calibrate the thermopile it is necessary to use as a source of energy some object whose radiant energy output is knowr, and constant. Coblentz ( 3 ) recommended thah for refined measurements of radiation stimuli a standardized lamp be used for calibrating purposes. Lamps previously made for standardization pGrposes emitted such a small quantity of radiant energy that the calibration points were below the range of readings obtained from a radiant type of gas heater. Consequently it was necessary to prepare special lamps for this purpose. W. E. Forsythe and his staff at the Neia Park Research Laboratory of the General Elec8 tric Company, Cleveland, Ohio, consented to make 7 and calibrate these lamps. Calibration of the lamps was accomplished by meas6 uring the mean spherical candlepower and then the 5 horizontal candlepower in the direction in which the lamps are used for standr4 ards. From the two reada & ings obtained the total enf’ ergy per square centimeter i in a definite direction was calculated. The r a d i a n t 2 energy of these lamps was given in watts per square , centimeter at a distance of 20 cm. from the plane of 0 the filament. 0b per S+Ft ~ B THr The standardized lamp Figure 3 was placed in a socket connected to 110-volt d. c. and the thermopile arranged to focus on the side of the bulb opposite the supports for the filament in a direction perpendicular to the plane of the filament directly in line with the marks on the side of the bulb and at a distance of 20 cm. To prevent reflection the table and walls were blackened. The line voltage was adjusted to 110 volts and the lamp allowed to heat thoroughly and readings as indicated in millivolts on the potentiometer were recorded: Readings obtained on one each of the following lamps-100, 300, 500, and 1000 watts-without the quartz window, are given in Table I.
Vol. 22, Yo. 6
ant curve is practically a straight line. It also indicates that an average arithmetical figure representing B. t. u. per square foot per hour per millivolt may be used. Table I1 shows the method of obtaining the mean value in B. t. u. per square foot per hour per millivolt. The mean value in B. t. u. per square foot per hour per millivolt is the figure used in computing the radiant efficiency of a radiant heater. Table 11-Method LAMP
SIZE ii’QtlS
of Calculating M e a n Value i n B. t. u. per Square Foot per Hour per Millivolt
THERMOPILE ENERGY AT U N I T ENERGY AT THERMOPILE THRRMOPILE READING E t u per sq ft ,Mzllrzolls B t u per sq fl p e r hr per mu
Method of Test and Calculation of Results
The A. G. A. Approval Requirements for Space Heaters (1930), in Part 11, Section 13b, specify that: “An appliance represented as a radiant heater shall have a radiant efficiency of not less than 25 per cent.” The method now used in this laboratory for the determination of radiant efficiency of space heaters consists of the use of a thermopile suitably mounted to obtain a sufficient number of readings to insure an accurate measure of the total useful radiant energy emitted. It was considered by the laboratory, after a study of the data at hand, that the effective radiant heat from a space heater was thrown out over the area determined by 130degree angles in the vertical and horizontal planes normal t o and through the center of the heater. It was also found that readings taken every 10 degrees gave results which are practically the same as readings every 5 degrees and could be made in about half the time. When the heater is set up for test, the centers of the radiants are placed exactly 2 feet (61 cm.) away from the face of the thermopile and at the center of the sphere over which the readings are taken. The heater is lined up by setting the thermopile a t the 0 degree position of the rail and the rail at the 0 degree position on the horizontal scale. The heater is then adjusted on the table until the center of the radiant is in line with the aperture in the thermopile and 24 inches (61 cm.) from its face. The rail is then swung through a 20-degree arc and a measurement made to the nearest part of the heater. The rail is then swung to the 20-degree position on the opposite side and a measurement taken at the same place on the opposite end of the heater. When these two 20-degree measurements are the same and the center of the heater is 24 inches (61 cm.) from the face of the thermopile, the heater is in the proper position and ready for test. The heater, which has previously been adjusted to the manufacturer’s B. t. u. rating, is then lighted, adjusted to normal pressure, and permitted to burn for ‘/zhour to become thoroughly heated and to reach a state of thermal equilibrium. The potentiometer is then balanced against its standard cell, after which the readings are taken. Table I-Readings Obtained in Calibration of T h e r m o p i l e The B. t. u. rate of the heater is first obtained and reLAMP LAMP POTENTIOMETER LAMP SIZE NUMBER READING VOLTAGEENERGY AT THERMOPILEQ corded. The rail is then swung to a 60-degree position and Watts p e r B. t. u . per clamped. The thermopile is clamped in successive 10-degree Wut1s Mdlioolts Volts sq. cm. so. ft. 41.1 0.4 110 0.0129 100 86-1676 positions along the rail from -t 60 to -60 degrees and themilli150.2 1.52 112 0.0475 300 86-1678 volt readings obtained for each position recorded. Care 109 0.0797 252.2 2.49 500 86-1675 796.0 7.65 110 0.251 1000 180 must be taken to make sure that the thermopile has remained a T o convert watts per square centimeter to B. t. u. per square foot per lofig enough in place to give a constant reading on the potenhour multiply b y 3170 (3.415 X 929). tiometer. Care must also be taken to prevent drafts from With millivolt readings as ordinates and B. t. u. per square playing across the heater and thermopile. The rail is then foot per hour as abscissas the calibration curve shown in Figure swung around 10 degrees and clamped. The thermopile is 3 was plotted. The curve shows that as the radiant energy moved along the rail and readings obtained as before. This increases the thermopile readings increase so that the result- process is repeated until the area over which the effective
INDUSTRIAI; AND ENGINEERING CHEMISTRY
June, 1930
599
radiant heat is thrown out is completely covered by readings at 10-degree intervals. This series of readings is taken with the fused-quartz window in place.
the weighted millivolt value and the area covered as well as the diathermacy constant. The formula for radiant efficiency then becomes
Correction for Diathermacy of Q u a r t z
D X I where M = weighted millivolt value A = area covered (16.451 square feet) K = thermopile constant (B. t. u. per square foot per hour per millivolt = 100 3 for thermopile used) D = diathermacy current I = input B. t. u. per hour R = radiant efficiency
Early in our experiments it was discovered that the quartzglass window used in the instrument did not transmit all the radiant energy. A11 solid objects, however transparent, absorb some of the radiant heat which falls upon them. Those which permit radiant heat to pass through them are said to be “diathernious.” The diathermacy of a substance usually depends upon the nature and temperature of the Pource of radiant heat. Since the temperature of the radiating elements varies with different heaters, the best substance to use for covering the slit would be one, if it could be found, whose diatherniacy was unaffected by this temperature variation. Among all substances tried, a polished disk of clear rock salt showed most nearly the desired characteristic. However, exposure to the intense radiated heat and moisture caused the disk to lose its polish and a change in reading would occur. I t was therefore decided to make a diathermacy correction for each heater using a quartz disk as the slit covering. This is done by removing the quartz window, setting the thermopile a t either the 10 or 0 position on the rail. swinging the rail to the 10-degree positions, and taking the millivolt readings for each position. This would give a series of readings across the face of the heater. The 10- or 0degree position of the thermopile on the rail is used, whichever gave the higher readings on the previous run. The sum of the first series of readings across the face of the heater divided by the sum of the readings with no window in the thermopile, and the thermopile in the same position as in the first series, will give a figure between 0.3 and 0.5, n-hich is called the “diathermacy constant.” The readings without the quartz window in place are taken with the slit in a vertical position and the slit is so constructed that the thermopile is unaffected by convection currents. To determine whether the ratio of the readings with the window in place to those d h the window out of place was constant, readings were taken a t the 0- and 10-degree horizontal positions on several heaters. The results on three of them are listed below and shorn that this figure is a constant within the accuracy of the measurements. HEATER 1 2 3
DIATHERMACY CO~STAET 00 10 0 413 0 418 0 355 0 349 0 389 0 381
+
R = MXAX K
SAMPLE CALCULATION
Step (1). Assume the following summations of horizontal values: WITHQUARTZ WINDOW +60° = 18 31 +50° = 22 70 +40° = 25 80 +30° = 26 56 +ZOO = 26 83 + l o 0 = 25 98 0’ = 23 28 -10’ = 19 84 -20’ = 14 78 -30’ = 8 84 -40’ = 4 09 - 5 0 0 = 1 33 -60’ = 0 55
65 40
Step ( 2 ) . The series of constants previously mentioned are calculated as shown below: hn H X Q where h, = height of a given zone represented by a thermopile reading H = total height Q = number of horizontal readings = 13 Factor
=
Factor at: 0 22408
50’ = 13
6252 = 0 004753 0 26706 40’ = 13 6252 = 0 005665 0 30192 30” = 13 6252 = 0 006406 0 32760 6252 = 0 006951 0 34332 100 = 13 6252 = 0 007284 oo = 0 34864 13 6252 = 0 007397
20’
= 13
Using the summations of step ( l ) , the weighted average of each horizontal series becomes: +60° +SO0 +40° +30°
Each reading of the thermopile represents the radiation emitted through a certain area, and since the area covered by each series of horizontal readings is smaller the farther away from the 0 position they are taken, it is necessary to correct them for the area actually represented. The matheniatical solution is not included here but the procedure for obtaining these factors is shown under step (2). To calculate the radiant efficiency proceed as follows: (1) Add all the millivolt readings across the sheet or, in other words, all the readings taken in the same horizontal plane. ( 2 ) Multiply these results by a series of constants to obtain a weighted average for the area represented in each horizontal plane. (3) Add the weighted averages so calculated. This gives a weighted average of the entire.series of millivolt readings. (4) Calculate the diathermacy constant by determining the ratio of the energy measured with the quartz window in place to the energy measured with no window in place. This value, = sum of horizontal readings of maximum value sum of no window readings taken over same arca’ ( 5 ) Calculate the radiant efficiency. Since from the thermopile calibration a value is obtained which is in B. t. u. per square foot per hour per millivolt, it is necessary to take into account
WITHOUT QUARTZ WINDOW
+zoo
+lo=
- 1;: - 200
-30’ -40° -50‘
- 60’
18 31 X 22 70 X 25 80 X 26 56 X
0 003698 004753
0 0 0 0 0 0 0 0 0
= 0 0676 = 0 1080 005665 = 0 1460 006406 0 . i698
26 83 X 006951 25 98 X 007284 23 28 X 007397 19 84 X 007284 14 78 X 006951 8 84 X 006406 4 09 x 0 005665 1 33 x 0 004753 0 5, x 0 003698
0.1869 0.1890 0.1720 0.1443 0.1026 0.0567 0.0231 0.0063 0 0023
Step ( 3 ) . Summation of values found in (2) = 1.3746 Step ( 4 ) . Calculation of diathermacy constant = Summation a t + l o o with window - 25.98 0.397 Summation a t + l o o without window 65.40 Steb (.5,) . Radiant efficiencv = 1.3746 X 100.3 16.451 X 100 - 25,0 per cent 0.397 X 23,100 23,100 = B. t. u. of gas consumed during test
.
+
General Discussion
By the above procedure 169 readings are taken, which permit a more correct average of the amount of heat radi-
ISDUSTRIAL A&\-D ESGISEERISG CHEMISTRY
600
'
ated than if observations are made a t only a few points. The area covered by these readings includes, we believe, all of the useful heat radiated from the appliance. Adjustment of the'instrument to the varying conditions may be made rapidly and the readings taken a fern seconds after the setting is made. All readings may be taken by one man in approximately 3 hours or by two men, one reading the potentiometer and the other changing the position of the thermopile. in approximately 1'/? hours. Accuracy of the method depends primarily on how accurately the potentiometer readings are made. The readings taken hare been found to vary froin 0 to 4.2 milli\-olts. the average over the whole area tested usually being about 1.5. Since readings on the potentiometer can be made within 0.02 millivolt, an accuracy of 1.3 per cent of the indicated value iiiay be expected. In other words. on the basis of potentiometer readings if an efficiency of 30 per cent is obtained, it is accurate to n-ithin *0.39 per cent. The entire apparatus and procedure are capable of giving duplicate results within this degree of accuracy as shown by Table 111.
Vol. 22. s o . G
monoxide, and ordinarily such a low percentage of air could not be obtained and still maintain complete combustion. (4) Increase the opening in front of the radiants by decreasing the area of the dress guard, setting the radiants well above the hearth plate and decreasing the amount or width of the overhanging back wall The dress guard is often made sc thick that it offers considerable obstruction to radiant heat, absorbing the radiant heat and giving it off principally as convected heat. By decreasing the size of the dress guard it is possible to increase radiant efficiency several per cent. Instances may be cited where the hearth plate was very wide and where the radiant efficiency was increased about 3 per cent by setting the radiants up above the main part of the hearth plate, thus permitting a greater angle for emission of radiation toward the floor. A wide overhanging back wall prevents radiant heat from getting away from the upper part of the heater. ( 5 ) The type and porosity of the clay seem to have a marked influence on the amount of radiant heat emitted. How much difference could be obtained between a porous clay radiant and a hard non-porous radiant has not been determined as yet, but apparently there is a marked difference in favor of the more porous radiant Possibilities of F u t u r e Research
That a n instrument of this type measures radiant heat only has been demonstrated by Coblentz (5) and Moll (12), but whether the condit,ions under which the thermopile is required to operate modifies its accuracy must be determined before it can be said that the results are entirely correct. Consequently, all heat except the radiant heat' from a radiant fire was determined as nearly as possible by determining the flue loss, other losses through t'he bottom, back, sides, and top of the heater, and subtracting from 100. From these determinations it appeared that the flue loss was approximately 45 per cent and other losses through the back, sides, t,op, and bottom were approximately 7.4 per cent. While it' is not claimed t,hat this procedure was extremely accurate, it shows that the radiant efficiency of 47.6 per cent was within a reasonable degree of being correct.
I n the preceding discussion very little has been said about improving the spectral quality of radiant heat. I t has been concerned chiefly with total radiant energy. While a reasonably high percentage of the gas burned is transformed into radiant heat in present-day heaters, it is believed that there is a possibility of making improvements whereby this percentage may be considerably increased. I n a recent article Kurokawa (11) claims that radiants coated with palladium, platinum, copper, iron, and oxide of uranium emitted more radiant energy than those without coating. Just how much effect such coatings would have is problematical, but it presents a n interesting vista for research. No attempt has been made, so far as the A. G. -1.Laboratory is concerned, to determine the percentage of the total possible of the various rays of the spectrum such as ultraviolet, infra-red, visible, etc., which are given off by a gas radiant heater, but it would seem to present a n interesting and possibly fruitful research problem. After having determined the amount of various kinds of rays present in the usual types of heaters, it should be possible to suggest ways of increasing the percentage of the short infra-red rays which seem to be desirable from a health and comfort standpoint.
How t o Increase R a d i a n t Efficiency
Acknowledgment
As a result of several years' experience in testing radiant heat,ers it has been found that there are many ways by which radiant efficiency may be increased. The time was not available to carry each possibility through to completion, so they are merely listed below. There may be other means of improvement, but those listed halve been proved to be the most common methods whereby radiant efficiency may be increased:
.Wmowledgment is here gratefully expressed for the assistance of all those contributing t o this paper and particularly to Messrs. Albrecht, Cook, Corsiglia, and Shawn.
of D u p l i c a t i n g R e s u l t s a n d R e l i a b i l i t y of M e t h o d of T e s t RADIANT EFFICIESCY Per cenl 1st t e s t . . . . . . . . . . . . . . . . . . . . . . . . 48.2 3rd test next d a y . . . . . . . . . . . . . . . . . 4 7 . 8 Average.. . . . . . . . . . . . . . . . . . . . . . . . 47.6 Max. variation between readings.. . . 0.4
T a b l e 111-Possibility
(1) Lengthen the tips inside of the radiants, thereby obtaining more heat-absorbing and radiating surface. ,4n increase of several per cent has been accomplished in this manner. (2) Increase the insulating effect of the back wall, thus permitting more heat to be thrown forward. The following data will indicate the effect of back-wall construction on one series of heaters, all other conditions being practically identical: Thick clay back wall Thin sectional clay back wall Sheet-metal back wall with one sheet of asbestos
P e r cent 32.; 31.28.1
(3) Decrease the free-air content of the flue products inside of the radiants. A heater which gave the highest radiant efficiency so far obtained a t the laboratory had only approximately 9 per cent free air in the flue products as sampled inside of the radiants a t the top. It might be said that this heater, being overadjusted, was producing a slight amount of carbon
Literature Cited Bone, Callendar, and Yates, J . G a s L i g h t i n g , 130, 72 (1913'. Callendar, Proc.. P h y s . Soc. London, 23, P t . 1 (1910). Coblentz, Bur. Standards, Scz. P a p e r 11, 87 (1914) Coblentz, Ihid., S c i . P a p e r 11, 131. Coblentz and Emmerson, I b i d . , Sci. P a p e r 261, 504 (1916' Coblentz, J . Optical Soc. A m . , 5, 3.56 (1922). Coblentz, Bur Standards, Sri. P a p e r 237 (1913). General Gas Light Co., Kalamazoo, Mich., "Radiant Heat for Health " General Gas Light Co., "When H e a t Is Health." Hartley, G a s J . , 185, 86 (March 13, 1929). Kurokawa, J . .SOL. Chem I n d , ( J a p a n ) , 32, 481 (1929): Sir.*?l Bi.rildi?ia, 32, 149 (1929). hloll, Proc. P h y s . Soc. London, 35, '757 (1923). Pacific Coast Gas Assocn, Gas Appliance Testing Code, p . 7 2 ( 1 9 N ) . Radiation, Ltd., London, "Radiant Heat and Comfort." Radiation, L t d . , G a s J . , .4dvertisement between pagei 17.' a n d 173, October 16, 1929. Ray-Glo Corp., Athens, Ohio, "Ultra-Violet Sun Heat in the Home." Reihr and Knoblauch, G a s W a s s e r f a c h , 69, 897 (1926). Schalek, Y a t u r a i G a s , October and Sovember, 1929. , 1927. Shawn, A m . G a s Assocn. M o n l h l ~ May, G a s J . , 160, 592.
